1868.] MAW VAIIIEGATED STRATA. 387 



to an ochreous yellow. Although these extremes are not very distinct, 

 and are connected by a series of gradations of intermediate shades 

 of colour, they are evidently due to independent causes. There 

 seems to be no uniform relation in the proportion of silicate of iron 

 in the light and dark parts of variegated beds ; and its amount seems 

 to vary independently of the action of discoloration. The proto- 

 silicate is probably the colouring principle of the blotches tending to 

 green. Bischof (Chemical and Physical Geology, vol. ii. p. 130, 

 English edition) also records the occurrence of this green proto- 

 silicate in variegated sandstone which had no green colour. In asso- 

 ciation with sesquioxide of iron, the green -colouring power of the 

 silicate would be obscured, but developed on the removal of the red 

 sesquioxide. This, we have seen, has taken place in every degree ; 

 and where the segregating motion of the red anhydrous sesquioxide 

 has occurred, it has generally become hydrous. The hydration of 

 itself, without any diminution of the sesquioxide, would produce a 

 lighter yellow blotch on the red ground ; indeed, in the red clays of 

 the Lower Bagshot beds, a yellow clay occurs, lighter in colour than 

 the red, though containing more of the sesquioxide of iron, and 

 rendered lighter simply by its becoming hydrous. 



The following are analyses of the red and yellow portions of the 

 variegated clay represented in fig. 23, PI. XIII. 



Analysis 'No. 33. Blood-red clay. Lower Bagshot beds. Messrs. 

 Pike's Clay Works, Wareham. (Dr. Voelcker.) 



Sesquioxide of iron with traces of protoxide . . . 5'28 per cent. 



Analysis No. 19. A second specimen of bright-red clay, Messrs. 

 Pike's Clay Works, Wareham, Lower Bagshot beds. (Laboratory 

 of the Museum of Practical Geology.) 



rp ^ , . r Soluble 206110^0 i. 



Total n-on | i^^^i^^le 0-459 l^'^^ percent. 



f Sesquioxide of iron 2-044 „ 



p . ,. J Protoxide of iron, soluble . . . traces, 

 j Oxides of iron, insoluble, 



(^ weighed as protoxide 0-590. ,, 



Analysis No. 20. Bright ochreous yellow clay, intermixed with 

 No. 19. Lower Bagshot beds, Messrs. Pike's Clay Works, Ware- 

 ham. (Laboratory of the Museum of Practical Geology.) 



m.,. /Soluble 5-3821 ;.„.^ 



Total iron | i^^^i^^le 0-361 ] ^'^^^^ 



f Sesquioxide of iron 7-69, 



p , j Protoxide, soluble a trace. 



present as ^ q^^^^ ^^ ^^^^ insoluble, 



( weighed as protoxide 0.4 65. 



The variegation of the Horderley (Caradoc) sandstone, fig. 28 (PI. 

 XIV.), in which green bands and blotches occur on a purple ground, 

 appears only accountable by the further hydration in patches of the 

 sesquioxide of iron to which the purple colour is due. 



The following determinations of the iron in the purple and green 

 portions fail to afi'ord any other explanation, as they contain iden- 

 tical amounts of sesquioxide. 



